In Capricorn’s Orbit

Since the day’s other article was a little on the short side, I felt like I’d treat to you to one of my sightings on a recent visit to Sweden.

The same year Honda dazzled us with the NSX and Lotus revealed the Carlton, BMW dazzled us with this mediocrity. I had forgotten that 1990 was such a special year. I don’t remember the press being very critical about the interior of this car. Compared to the E30 this is shocking, the kind of thing BMW drivers thought that Opel and Ford drivers had to endure. Having worked out that their customers never went near the Blue Oval or the GMpire, BMW decided this kind of thing would do just fine.

Did the Ulm-educated designers take early retirement at BMW?

This is BMW’s attempt at an airbag embedded in the dashtop. The panel gaps don’t quite line up with those on the inexplicably busy ventilation panel below it – notice the feature line running sideways from vent to vent. That padded panel has two different radii in the corners.

The devil is in the details

I really never expected these discoveries. The next thing is to take a look at the interiors of the cars on sale circa 1994 to see how to place this nasty confection.

Author: richard herriott

I like anchovies. I dislike post-war town planning.

27 thoughts on “In Capricorn’s Orbit”

  1. Contrary to popular belief BMW was not particularly awash with money at the time the E36 was developed. That car’s new Z axle rear suspension was quite expensive to make and BMW still couldn’t ask the prices customers were willing to pay for their Benz 190 or C Classes so they had to save some money and they decided it was the E36’s product quality that had to create the earnings.
    The car in the pictures is from the second production run with a grained plastic insert atop the glovebox lid. The first cars had a strip of shiny black plastic there and were even worse in terms of material quality and particularly rust protection.
    The interior was crudely made with exposed screws everywhere and it had sometimes ridiculous material quality like at the ventilation outlets. BMW also thought that plastic inner door handles were sufficient – interestingly the better equipment levels for UK vehicles had chromed items that weren’t available anywhere else. If the customer was unhappy enough to have paid through his nose for electric window winders he found switches that got so hot he could burn his fingers before their illumination went LED late in the model’s life.
    The E36 was half ready when it entered production resulting in innumerable detail changes to reduce wind noise at the most improbable places (remember the headlamp glass covers of the second series with their dimples at the top edge to create a turbulent air flow and prevent whistling noises at the gap to the bonnet? How many different rear view mirrors did they use all in vain attempts to reduce wind noise at the A post?) and it was badly made. Paint with orange peel on the roof was standard, the integral door window frames of the saloon’s rear doors never could be made to acceptable tolerances.
    The E36’s biggest faults surely were its contortionist seating position (at least in LHD cars) and its complete lack of any directional stability on uneven roads which in combination with the mushy steering of the later cars made it impossible to hold the car on a straight line.

  2. I owned an E30 325i convertible when the E36 was launched and remember thinking how unfavourably the E36 compared with its predecessor, particularly the very “plasticy” interior. Externally, it was certainly a more modern design, but lacked elegance. The one-piece door pressings with their thick window frames and wide shut lines put me in mind of the Sierra. Dave’s insight about wind noise and production tolerance issues doesn’t surprise me.

    One issue in particular defined it’s (relative) cheapness: the integral plastic bumpers on the saloons were unpainted grey plastic when colour matched items were already the norm, even on mainstream models (and the facelifted E30 had them). BMW defended this by arguing that they were easier to recycle, but when the two-door E36 was launched, the bumpers were painted and colour matched. The saloon gained the same at its first facelift.

    1. Very little of those weaknesses were discussed in the reports of the day. I mean, I don´t remember the writers commenting all that much. In its base-model spec it was a deeply ordinary car.

    2. I had a look at some of the mid 80s cars BMW may have benchmarked and they all look neater and tidier in terms of assembly concepts. You can argue a bit with the style but not the way they were designed to be assembled. The Sierra and Renault 21 are very neat indeed.

  3. All this seems a bit harsh.

    Sure, the E36 might have been badly made with very questionable material quality, but it was hardly alone in that at the time. The Golf Mk 3 had pretty poor quality at launch too, and it was still a cut above most European contemporaries.

    I’ve always considered the E36 a brave and smart modernist design… perhaps they changed too much all at once, without the required budget to do it properly?

    1. Most of this car’s peers were quite alright in terms of assembly concepts. The Golf Mk3 – was that a Lopez cost-cut victim? This BMW might be an intermediary car. The next one had the kind of assembly solutions one would expect of BMW. It was no charmer but it withstood a detailed inspection. A family member had one so I have seen it up close.

    2. I think you might be right about the degree of change on a limited budget conundrum, Jacomo. I’ve always though of the E46 3-Series as the car the E36 should have been: it took a second attempt (and more money) to get it right. BMW used to articulate the line that their model succession strategy was to alternate between evolution and revolution. Hence, the E30 was an evolution of the E21, the E36 was a clean-sheet design that evolved into the E46. Of course, that all broke down in the Bangle era.

      Interestingly, BMW reverted to more traditional (elegant?) separate and slimmer door window frames for the E46. As far as I am aware, the E36 was the only BMW ever to use the single-piece door skin pressing.*

      *Richard, are you able to enlighten me as to the correct technical term to describe these one-piece door pressings?

    3. “one-piece door pressing” should suffice. Do you mean the exterior door or interior door? The exterior door is, if I recall correctly, at least two peices with welds at the shoulder-line where the door becomes the door frame.
      Bangle disrupted the design evolution of BMW not the engineering evolution. I don´t know which of his cars were visually and mechanically different from their predecessor versus only visually different.

    4. The E36’s door skins are one piece pressings (with lower halves available for repairs), as were those of all Golfs up to Mk4 and many other cars. The Golf used these doors because the Beetle had them as it was designed around Beetle production processes to use the existing production equipment.
      These doors are very difficult to produce with tight tolerances because window frames can’t be moved independently form the door body to compensate for variations. When BMW introduced the E36 their competitors (particularly in Ingolstadt) shook their heads because it was quite clear that BMW had no experience with this kind of design. I’ve never seen an E36 saloon where the rear upper edge of the door properly fits in with C post and roof panel, this corner of the door always stands proud for a couple of millimetres and the panel gap is very uneven.

      BMW had to build up amortization reserves of three digit million Deutschmarks to buy back E36s from the first production run to make sure nobody saw a rusty BMW on the road…

    5. I often wondered about that: is the hole for the window cut out of the sheet prior to pressing? I had considered that option but suspected it would result in a lot of wasted steel bits being left over.

    6. @Richard: please look at the part in the following picture and imagine the amount of steel cut out/off and thrown away.

      And all this with a material that largely behaves like a leaf spring and will jump back somewhat from the form it was stamped into.

      As long as Schuler can provide the necessary hydraulic press tools and CAD support correctly calculates spring rates and creep factors of the steel it’s easier, cheaper and more precise to make such large parts. Quite different from PSF’s approach of stitching together myriads of steel snippets.

    7. Dave: thanks for that. The image explains a lot. I´m not from an engineering background and more of a plastics than steel person. I had no idea they were pressing such large items. There were a number of cars whose construction I could not decide upon because I was thinking of how smaller bits could be put together rather than of really large items. I presume there are advantages and disadvantages to larger pressings. I´d be fascinated to know how much of those factors work their way back to the modelling of the appearance stage. Some of it must be intuitive yet there must be cases where unmakeable details worm past the initial checks. I will check up on this Schuler company. Thanks for the detail!

    8. Dave: I went and had a preliminary look at Schuler´s website. I was almost giddy with excitement. Those machines are stunning. All of a sudden my conception of steel pressing moves from about 1956 sharply forward to 2018 in one huge leap. They have these huge enclosed machines which, I guess, remove the chance of injury and reduce noise and also can handle the parts from start to finish, including the bit where the part has to be handled and stacked.

    9. I visited Skoda’s new stamping plant at Mlada Boleslav last year, which was an impressive experience.

      You, Richard, won’t be surprised when I tell you that the stamping line (by Schuler, incidentally) itself is as big as a block of flats. What may be more surprising is the information that it’s not terribly noisy during operation – and that the current generation of stamping machines facilitates the exchanging of dies in less than five minutes. All of which I found terribly impressive – just like the price of an up-to-date Schuler press (€80 million).

    10. Kris: it gets better. For some of their tooling machines, the tool switch over is three minutes. The noise suppresion is a big part of these machines value. Noise eventually harms people.

  4. Ah, right. I thought there was a technical term to describe doors where the external skin and window frame is (or appears to be) pressed from a single piece of steel, rather than have a separate and distinct window frame.

    Regarding the revolution / evolution strategy, I was referring to design, not engineering. The E28 5-Series appeared, at first glance, to be a light facelift of the 1972 E12 original, yet shared no bodywork in common. The E34 represented the same revolution as the E36 did for the 3-Series, and evolved into the E39 fourth generation 5-Series. Had this strategy continued, then Bangle’s controversial E60 5-series should have been evolved into its successor. Instead, they lost their nerve and started with a clean sheet to produce the rather underwhelming F10 and this has been (undeservedly?) evolved into the current G20 model.

    1. Daniel,

      Agreed, I’ll happily pile on the F10 5 series because it’s easily the worst of the breed.

      I always thought the change of heart was because of demand from developing markets for ‘more luxury’. So BMW based the F10 on a cut down 7 series platform and said ‘will this do?’

      It sold well but was disappointing from an engineering and design point of view.

  5. Talking of dodgy BMW production engineering, wasn’t there a 7-series that didn’t have symmetrical wings or something because of a mistake in the clay model? Or is that an urban myth? Surely you’d just take the CAD/CAM measurements of the clay model for half the car and compute the other half to ensure perfect symmetry…

    1. The E32 Seven was asymmetrical to the extent of several millimetres at the front wings as was every other car from that time. Design work on the E32 was done primarily in clay and later transferred to digital data and clay models never are completely symmetrical (today the design is created in CAD and only checked in 3D clay as a verification measure probably the most prominent reason for the poor designs of today).
      As hardware performance was nothing like today and algorithms for 3D data processing were very rudimentary at that time there was no possibility of making the data symmetrical after transferring it from clay to CAD. That was not limited to BMW but everybody else had the same problems.
      The E38 was the first BMW with absolutely symmetrical bodypanels because CAD data were mirror imaged.

  6. I posted quite a lengthy comment yesterday in reply to Richard and Dave that seems to have got stuck in the moderation queue. Could an admin take a look please?

  7. Leaving aside the quality issues, the E36 had a big impact on my teen self: it was at the time, for me, a refreshing proposition, looked modern like no BMWs had before and put BMW on the map in my mind.
    It probably helped its case that the sister of my then best friend bought herself one pretty much right after launch, I could admire it almost everyday, it was a kind of Burgundy colour and didn’t have the base all-grey plastic bumpers.

    Even then, I think I could notice the E36’s seemed more ‘plasticky’ and less ‘metallic’ than its predecessor but it did work in its favour for me then, the ‘Metallic” older E30 looked very 70s/80s and passé while the more ‘plasticky’ E36 looked thouroughly modern and bang-on for the times.

    Today I understand that the E30 was just better built than its replacement and I came to appreciate well-executed bright works and top quality execution but despite its shortcomings I think the E36 did have a positive impact of sorts at the time, at least in some minds !

    1. …..and I think that -Age- has a big impact on how we appreciate a design or not. I think, sometimes, we tend to idealise things that we discovered between say the age of 10 and 18 and then hold everything else against that idealised teen version: in that sense Iam not surprised the E36 had a bigger impact on me than the E30.
      I think it could explain for example why some hard-core Citroenistes hold on to a CX-era image of Citroën that, of course, will never come back. They all tend to be around the same age-bracket lol.

    2. Indeed – around the early twenties one has enough experience to understand the world around one and also one assumes that the world then is the authentic thing. Later changes can seem inauthentic and a deviation from a personal norm that is mosly an accident of birth. Carwise, the period of “learned normal” is 1990-2000. However, one can also gain a wider perspective and re-assess the found world. My view of urban planning is expansive enough to see the period 1945-1995 as one of profound disruption. In 1990 I had no such image: all of it was lumped together in one mass and I think I was very much more accepting of Modernism than I am now. In Ireland the framing was Modernism versus low-grade pastiche vernacular – no competition. Seen agains the broad sweep Modernism´s tenets and implementation seem very wanting. So, I prefer design from centuries before I was born. Some new architecture is first rate but there is not a lot of it around.

    3. Thank you for the hindsight Richard. I read too a while ago that the reason the memories of our late teens/early twenties have the biggest impact and remain the strongest throughout life is also because this is the time when the brain is at its best, the neuroplasticity working at its optimal level. Because the brain is then able to make the strongest connections it stores more vividly the memories and emotions of that time of our lives. Hard to explain but fascinating stuff for me nonetheless.

    4. Up until the industrial revolution, life had a slow pace. If your perception of normal was set in your twenties by the time you died at 45 little would have changed. Thinking back to 1989 computers were either office-bound PCs or almost useless Commodore 64s. People had pagers and a very few had mobile telephones and people still read books and newspapers. Since then the pace has been intensifying until we have one day news cycles and there´s a new BMW launched every three months. No wonder some want to withdraw to a sun-dappled Tuscan terrace where it is always eight thirty in the evening.

  8. This is the comment I posted that WordPress helpfully swallowed. The moment’s probably passed now.

    Richard: I always thought the hole for the window was cut out of the sheet after the pressing, the simple reason being that’s what it looks like from the panel van variants of many cars, where it appears the metal panel where the window goes is simply not stamped out:

    Dave: Thanks for the detail about the E32 asymmetry. I write computer software for a living, I’m not an automotive engineer. With that huge disclaimer out the way, I’m surprised to learn it wasn’t possible to make the data symmetrical back in the day, even given the limitations of the computer hardware and software car manufacturers had access to in the 1980s.

    My understanding of how it worked is that the clay model was produced by incredibly skilled people and then a probe is painstakingly run over its surface to take the thousands of measurements that form the computer model of the car (there’s a good example of the process here: This gives you thousands of co-ordinates in three dimensional space.

    To my simple mind you’d measure along the centreline of the car (in plan view) to establish an origin and then you’d only need to change one axis for each co-ordinate to flip the handedness. For example, if you have a measurement for a point where the front wing meets the edge of the bonnet, you wouldn’t need to change the height or how far back it is from the front of the car, only which side it’s on in relation to the origin. Note that I avoided using x, y and z in that last sentence because I wasn’t sure which way up it goes for modelling cars. Like I said, I’m not an engineer so there’s probably something really significant I’m missing. I’m sure if they could have made the data symmetrical they would have!

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

This site uses Akismet to reduce spam. Learn how your comment data is processed.

%d bloggers like this: